SEARCH

SEARCH BY CITATION

Keywords:

  • Africa;
  • apes;
  • behavior;
  • disease ecology;
  • disturbance;
  • ground use;
  • habituation

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The potential of human activities, including research, to alter parasite transmission ecology in wildlife is unknown. We examined gastrointestinal parasitism in chimpanzees (Pan troglodytes schweinfurthii) in Budongo Forest, Uganda. Trail use and time spent on the ground was recorded during 10 months of observations in four sites with differing human disturbance. Disturbance was quantified using transect plots (n = 320). Fecal (n = 435) samples were examined for helminth eggs, larvae, and for protozoan cysts. Individuals that spent more time on the ground had more infections and higher intensity infections. Prevalence of 13 parasite species was similar across sites, but percentage of multiple infections and infection intensity differed, as did ground use. Chimpanzees at the long-term research site spent more time on the ground or on human trails. We hypothesize that researcher presence and trail creation may influence ground use, and thereby parasite burden, by altering trade-offs between foraging and predation risk.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Researcher presence in protected areas can benefit conservation by reducing poaching pressure, generating employment, and raising public awareness (Campell et al. 2011). Research stations across 109 African resource management areas positively impacted great ape survival (Tranquilli et al. 2012).

However, elsewhere, adverse effects of research activities on wildlife have been reported. Researcher handling of incubating short-tailed shearwaters (Puffinus tenuitostris) reduces hatching success by up to 100% (Carey 2011). Habituation and surveillance alter activity budgets and social behavior of some animals (Newsome et al. 2005). Gorillas (Gorilla gorilla gorilla) decrease feeding rates with increasing researcher team size (Klailova et al. 2010). Moose (Alces alces) respond to research helicopters by increasing movement rates (Stoen et al. 2010). The use of telemetry, biologgers, and marking techniques, including colored tags, increases energetic costs, influences breeding ecology and susceptibility to predation in several species (Calvo & Furness 1992; McMahon et al. 2011).

Wildlife disease ecology may also be affected by research activities. Feeding of chimpanzees (Pan troglodytes schweinfurthii) at Gombe National Park is positively correlated with respiratory illness rates during the early dry season (Lonsdorf 2011). Transmission of pathogenic respiratory viruses from researchers to wild apes has been documented in Cote d’Ivoire, Rwanda, and Tanzania (Kaur et al. 2008; Köndgen et al. 2008; Palacios et al. 2011).

Researchers may exert subtle effects on wildlife by altering habitat. The creation of trails is common at long-term research sites (Strier 2010). The impact of these activities on animal behavior or disease ecology is unknown. Habitat alteration can affect patterns of parasitic and bacterial infections within primate populations (Gillespie et al. 2005a; Gillespie & Chapman 2006, 2008; Goldberg et al. 2008). Primates in logged forest (i.e., Cercopithecus ascanius [Gillespie et al. 2005b]; Alouatta palliata [Stoner 1996]) have higher parasite prevalence and diversity than individuals in continuous forest. Although the exact mechanisms through which habitat alteration affects disease transmission needs elaboration, it is plausible that forest degradation results in increased ground use and associated increases in infection risk (Nunn et al. 2000; Nunn & Alitzer 2006). Nematodes, such as Strongyloides spp., may be acquired from soil (Cheng 1986). If primates use the ground more frequently as a result of research activities such as trail creation they may be vulnerable to terrestrial pathogens. This is of conservation significance, as although intestinal parasites have more subtle effects on survival and fecundity than do acute viral infections, their cumulative impact may be similar (Gillespie 2006; Nunn & Alitzer 2006, Howells et al. 2010). To shed light on these issues, we examined ground use and patterns of parasitism in four communities of chimpanzees (P. t. schweinfurthii), living in one forest, each experiencing different levels of research activity and human disturbance.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Study site

Budongo Forest Reserve (435 km2) lies on an escarpment of the Western Rift Valley (between 1°37’N–2°03’N and 31°22’E–31°46’E) in Uganda and consists of mid-altitude moist semideciduous tropical forest (Reynolds 2005). The area experiences a bimodal pattern of rainfall, with peaks between March and May, and September and November (Reynolds 2005). Mean annual rainfall is 1,600 mm with daily temperature maxima and minima between 32°C and 19°C (Reynolds 2005). Four sites were selected for study, approximately 10–35 km apart. Each site has different history of human activity and a separate chimpanzee community (Table 1).

Table 1. Details about disturbance history and chimpanzee community at four study sites in Budongo Forest, Uganda
 SonsoBusingiroKaniyo-PabidiKasokwa
  1. a

    During study period.

  2. b

    (Sugiyama 1968; Reynolds & Reynolds 1965).

Approximate1°43′556′′N,1°42′941′′N,1° 55′099′′N,1°39′474′′N,
Location31°32′827′′E31°28′201′′E31°43′153′′E31°38′880′E
DisturbanceResearcher presence since 1991Briefly used for research in 1960sEcotourist SiteChimpanzees monitored since 2000
 Logged 1947–1952Logged 1935, 1980–1992Never loggedForest Fragment
 Site of old sawmillSignificant illegal logging and snaring  
Chimpanzee community sizea67Unknown, perhaps 80b7816
Habituation levelWell habituatedNot habituatedLess than Sonso, more than KasokwaLess than Sonso, less than Kaniyo-Pabidi

Disturbance plots

To quantify human disturbance, four 1-km transects were walked at each site. Transects ran parallel to each other, 500 m apart. At 50 m intervals along the transect, 12.7 m radius circular plots were marked (n = 80 plots/site). All forms of human activity were quantified, including snares, tree stumps, pit-sawing (logging) sites, evidence of pole collection, charcoal burning, and trails created by researchers or used for ecotourism. Each transect was walked once between October 2007 and April 2008. Data collection alternated between sites to minimize any effects of seasonal differences in disturbance. The first transect was completed at each site before the second was started. Sampling differed slightly at Kasokwa. Due to the small size of the forest, it was not possible to create 1-km transects. Rather, plots were marked along existing paths in the forest.

Feces collection

Behavioral data were collected at each site concurrently with the disturbance data, between August 2007 and May 2008, a period covering both rainy and dry seasons. Chimpanzees were followed from dawn until dusk. Parties were located by calls, a focal animal chosen and tracked for at least 1 hour. Afterwards a new focal individual was selected. When choosing focal animals, preference was given to individuals that had not been sampled within the previous 48 hours. During the observation period, the focal's behavior, including vertical height, was recorded every 5 minutes. Vegetation type was recorded at each time point, as: (1) Human disturbed habitat (agricultural fields, research camps, or sawmills); (2) Forest, canopy with open patches; (3) Forest, closed canopy; (4) Woodland (colonizing forest with savannah and forest species, but Maesopsis eminii scarce); (5) Human trails; and (6) Roads.

Due to lower levels of habituation, it was not possible to conduct focal follows of Busingiro chimpanzees. Behavioral data are therefore reported only for Sonso, Kasokwa, and Kaniyo-Pabidi. Most focal observations were made between 8:00 am and 13:00 pm (63.3% Sonso, 71.5% Kaniyo Papidi; 76.2% Kasokwa). Both fecal samples and behavioral observations were available from a total of 70 individuals (31 at Sonso, 28 at Kaniyo-Pabidi, and 11 at Kasokwa).

Fecal samples were collected opportunistically from chimpanzees at all four sites. All feces collected were freshly voided by individuals and stored in 10% buffered formalin. Care was taken to avoid collecting soil, foliage, or standing water contaminants. The tube was shaken to mix the sample with solution, labeled with date, location, and animal identifier. Samples were stored at room temperature until shipped to Emory University under appropriate regulations.

Parasite analysis

We examined 435 fecal samples for helminth eggs and larvae and protozoan cysts using concentration by sodium nitrate flotation and fecal sedimentation (Gillespie 2006). Slides prepared by each method were examined with a compound microscope. Parasites were identified on the basis of egg, larvae, or cyst coloration, shape, contents, and size. Each parasite species per sample was quantified and representatives measured at 400× to the nearest 0.1 μm with an ocular micrometer. Slide scans for protozoans were performed using the 40× objective. If needed, one drop of Lugol's iodine solution was added as a stain to facilitate identification. Unknown and representative parasite species were also photographed for later identification. Two slides were examined per sample for sedimentations.

Data analysis

Patterns of parasitism are described in terms of prevalence of infection, taxon richness, and the magnitude of multiple infections, fecal egg count and, in the case of protozoa, the intensity of infection. For aggregate parasite groups (nematodes, ciliates, and total infections), individual averages were analyzed, hence observations were statistically independent. The “total infections” variable excluded ciliates thought to be beneficial to an individuals’ health, of which the genus Troglodytella was recorded here. Analyses specific to ciliates included all ciliates.

Prevalence of individual taxa was defined as the proportion of individual samples infected with a particular parasite. Some individual animals contributed numerous observations, whereas others contributed only one—the degree of this skew prevented the use of individual ID as a random factor in mixed models. As repeated sampling of unknown individuals likely occurred at some sites, results presented here should be viewed as an “index of prevalence” (Gillespie 2006; Gillespie et al. 2008). Richness is defined as the number of unique intestinal species documented in the fecal sample (Muehlenebein 2005). Intensity is defined as the number of adult individuals of a parasite in a particular host (Margolis et al. 1982). We report intensity of infection for protozoan parasites. For other parasites, we can only report intensity of egg count in feces. We assume this is a biologically meaningful indicator of infection although it cannot provide a reliable index of adult worm burden (Gillespie 2006). For ciliates, where counts were not often complete (e.g., 500+), an index of total intensity was derived based on 20% percentile points of nonzero counts, giving a scale from 1 to 6.

Prevalence of individual parasite taxa was compared using logistic regressions effected with the SAS GENMOD procedure applying binary errors and a logit link function. Fecal egg count was, as is common, frequently highly aggregated (overdispersed) so that normal error models did not fit well. Hypotheses concerning intensities and egg counts were therefore tested using Generalized Linear Models correcting for overdispersion (SAS GENMOD procedure). ANOVAs on log-transformed parasite counts were used to examine differences in fecal egg count for parasite genera.

The relationship between behavior and parasitism was examined by comparing the proportion of time spent on the ground, and the mean observed scan height, with parasite richness, and the magnitude of multiple infections. Focal observations of chimpanzees included an estimate of the vertical distance from ground level for each individual at each time. These estimates were used as an ordinal scale in 5 m increments: 0 = on ground, 1 = 0–5m, 2 = 5–10 m, up to 9 = 40m +. This generated a data set of 10,506 observations, unevenly distributed among individuals. Later, we attempted to disentangle the effect of different predictors by fitting models including both proportion of scans on the ground and mean height of other (nonground) scans as predictors. In these models, mean height was calculated for all the observations not at h = 0 and entered to sequential (type 1) models after P_ground. The amount of time spent in different vegetation types, including human trails, was compared between sites, using similar techniques to White (1992).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Vertical height and ground use

The height at which chimps were observed differed between sites and between hours throughout the day (by site F(2,133) = 21.9, P < 0.01 and by hour F(12,638) = 4.8, P < 0.001). Chimpanzees at Sonso tend to spend more time at lower heights than chimpanzees at Kasokwa or Kaniyo-Papidi. At all three sites, chimpanzees tend to move down to lower heights over the course of the day (Figure 1A). There was no evidence that the temporal pattern differed according to site (site*hour interaction F12,638 = 1.4, P = 0.13). The overall proportion of time individuals spend on the ground also differed significantly between sites and hours (F2,133 = 24.07, P < 0.0001, and F12,638 = 4.29, P < 0. 0001). Chimpanzees at Sonso spend a greater proportion of time throughout most of the day near the ground than do chimpanzees at the other two sites (Figure 1B, Tukey Post Hoc Test).

image

Figure 1. (A) Mean height class over all observations by site. Error bars indicate standard error. Solid line = Kasokwa; Long dash = Kaniyo-Pabidi; Short dash = Sonso. (B) Proportion of time spent on ground over all observations by site.

Download figure to PowerPoint

Trail use by chimpanzees differed significantly between sites (F2, 98 = 24.59, P < 0.0001). Chimpanzees at Sonso were seen on human trails more often than were chimpanzees at Kaniyo Papidi or Kasokwa (Table 2).

Table 2. Percentage of time chimpanzees seen in different types of vegetation at three different sites
Vegetation typeSonsoKaniyo-PabidiKasokwa
Human disturbed18.6  0 1.7
Open with thick undergrowth 4.6 1.818.3
Forest, Canopy with open patches60.320.9 53
Forest, Dense canopy 9.274.8 6.6
Forest edge  2 1.420.3
Road 0.5  0 0.1
Trail 4.9 0.6  0
Woodland  0 0.5  0

Parasite infection

Nearly all samples (99.31%) contained at least one parasite identified to genus level. Thirteen parasite species were recovered, including eight nematodes (Oesophagostomum sp., Necator sp., Probstmayria sp., Strongyloides fulleborni, Trichostrongylus sp., Ascaris sp., Trichuris sp., and an unidentified strongyle), one cestode (Bertiella sp.), one trematode (Dicrocoeliidae sp.), and three protozoa (Troglodytella sp., Troglocorys cava, and Balantidium coli). Nematodes and ciliates were the dominant infection types, in 96.8% and 94.7% of samples respectively.

Effect of ground use on infection

Individuals seen on the ground more frequently tended to have both more ciliate and nematode infections, and higher intensity of infection, than other individuals within the same site (Table 3). There was no evidence that these effects differed between sites, or that there was a difference between sex (Site*P_ground interactions and sex effects were nonsignificant, and are not tabulated). At Sonso, both total infection richness and overall nematode richness recorded in individual chimpanzees increased with increasing frequency of observations on the ground (Figure 2A). When individual nematode taxa were examined separately (also using means for individual chimps as data), there was strong evidence for an effect of height on Strongyloides (F1,65 = 19.95, P < 0.001), but no evidence for an effect on Trichuris (F1,65 = 0.10, P = 0.75). Trends remained similar when the mean height of observations was explored as a predictor of infection patterns (Figure 2B). While frequency of observation on the ground and mean height predictors were correlated (r = –0.37, P < 0.001), individuals that were on average closer to the ground had higher total infections after adjusting for their tendency to spend time on the ground (F1,63 = 3.84, P = 0.054, Figure 2C shows the Sonso pattern). For other responses, the effects could not be disentangled.

Table 3. Effect of site, mean observed height, and proportion of time spend on the ground (Ground_P) on ciliate and nematode intensity and number of infections. GLM normal errors models (all error dof = 66). Responses were averaged across samples for each individual. Parameter estimates for Mean_Height all negative, and positive for Ground_P. All responses square root transformed, with the exception of ciliate infection number
  DFType 1 SSMean squareF-valuePr > F
Ciliate intensitySite213.53 6.77 6.000.0040
 Mean_Height116.7416.7414.840.0003
No. ciliate infectionsSite2 0.28 0.09 5.070.0089
 Mean_Height1 0.02 0.02 1.120.2938
Ciliate intensitySite213.53 6.77 5.300.0073
 Ground_P1 7.00 7.005.40.0222
No. ciliate infectionsSite2 0.28 0.09 2.250.1140
 Ground_P1  0.023  0.023 3.960.0509
Nematode fecal egg count intensitySite218.34 9.17 3.200.0470
 Mean Height111.0711.07 3.870.0535
No. nematode infectionsSite2  0.829  0.414 4.940.0100
 Mean Height1  0.816  0.816 9.740.0027
Nematode fecal egg count intensitySite218.34 9.17 3.140.0497
 Ground_P1 7.40 7.40 2.540.1160
No. nematode infectionsSite2  0.829  0.414 4.810.0112
 Ground_P1  0.661  0.661 7.680.0072
image

Figure 2. (A) Number of infections recorded in individual Sonso chimpanzees (mean among samples) and proportion of scans on the ground (F1,29 = 11.48, P = 0.0020. Without the two individuals at 1.0 on the x-axis, F1,27 = 1.83, P = 0.078). Each point represents a single chimp (average among samples where sample number for a chimp >1). (B) Mean number of nematode infections recorded and mean height category at Sonso (F1,29 = 8.48, P = 0.0068). When two observations between 0.0 and 1.0 on the x-axis are removed F1,27 = 1.24, P = 0.27. (C) The relationship between residual total infection number (eliminating P_ground effect) and mean observed height.

Download figure to PowerPoint

Chimpanzees that we observed spending more time on trails had, on average, higher mean ciliate infection indices (Figure 3, Pearson r = 0.55, P < 0.001). There was no evidence for an effect on nematodes.

image

Figure 3. Percentage of scans in which an individual was seen on trails and mean ciliate infection index at Sonso (solid) and Kaniyo-Pabidi (open) (r = 0.53, P < 0.001).

Download figure to PowerPoint

Infection patterns between sites

Intersite parasite patterns tended to match the within-site effects, with parasite burdens higher in sites where individuals were on average closer to the ground. Ciliate burden and infection intensity were significantly higher at Sonso compared with Kasokwa (Tukey means separation, P < 0.05). The mean ciliate infection index at Sonso was 4.15 (SE = 0.20) compared to 3.31 (0.21) at Kaniyo Papidi and 3.17 (0.46) at Kasokwa. Nematode egg count intensity was highest at Kasokwa 22.98 (8.25) compared with 16.45 (2.22) at Sonso and 11.43 (3.61) at Kaniyo Papidi (Tukey, P < 0.05, for Kasokwa vs. Kaniyo Papidi). The mean number of nematode infections was highest at Sonso (2.39, 0.15), intermediate at Kasokwa (1.97, 0.22), and lowest at Kaniyo Papidi (1.80, 0.16), with Sonso significantly higher than Kaniyo Papidi (Tukey, P < 0.05).

When data were examined by genus, significant differences in parasite prevalence existed between sites in Oesophagostomum, Trichostrongylus, Strongyloides, T. cava, and Troglodytella (Table 4), though directionality was not uniform. The percentage of multiple infections was greatest at Sonso (78.6%) and Busingiro (75.2%), even when multiple samples from the same individual were controlled for (83.3% Busingiro, 54.5% Kasokwa, 69.4% Kaniyo Papidi, 93.5% Sonso). The mean number of infections per sample differed significantly among sites [F3,431 = 4.15, P = 0.0017, Busingiro mean (SE) = 5.15 (0.16), Kasokwa mean (SE) = 4.41 (0.24), Kaniyo Papidi mean (SE) = 4.46 (0.148), Sonso mean (SE) = 4.92 (0.14)]. Infection intensity of T. cava and B. coli differed between sites, both being highest at Sonso (Table 5).

Table 4. Parasite prevalence % by site. Standard errors in parentheses. Multiple samples from individuals were included in the calculations. Differences in infection prevalence between sites were compared using SAS GENMOD logistic regression models including only “site”
Taxon Sonso (n = 173)Busingiro (n = 113)KP (n = 96)Kasokwa (n = 53)GENMOD χ2, P (3 dof)
NematodaOesophagostomum sp.82.7 (2.9)70.8 (4.3)82.3 (4.0)58.5 (6.8)15.5 (0.0014)
 Necator sp.37.6 (3.7)44.2 (4.7)31.3 (4.8)39.6 (6.8)3.24 (0.36)
 Trichostrongylus sp.36.4 (3.7)54.9 (4.7)39.6 (5.0)35.8 (6.7) 11.09 (0.0113)
 Probstmayria sp.14.5 (2.7)10.6 (2.9) 7.3 (2.7) 7.5 (3.7)4.18 (0.25)
 Strongyloides fulleborni40.5 (3.7)49.6 (4.7)26.0 (4.5)60.4 (6.8)  20.62 (<0.0001)
 Trichuris sp.1.2 (0.8) 5.3 (1.2) 1.0 (1.0) 3.8 (2.6)5.75 (0.12)
 Ascaris sp.0.6 (0.6)0.0 (-)  0.0 (-)  0.0 (-)  1.85 (0.61)
 Unid. Strongyle9.8 (2.3)11.5 (3.0)9.4 (3.0)20.8 (5.6)4.16 (0.19)
ProtozoaUnidentified20.2 (3.1)18.6 (3.7)25.0 (4.4)13.2 (4.7)3.26 (0.36)
CiliatesTroglodytella sp.96.0 (1.5)92.0 (2.6)96.9 (1.8)84.9 (5.0)9.28 (0.03)
 Balantidium coli0.0 (-)  0.0 (-)  0.0 (-)   1.9 (1.9)4.23 (0.24)
 Troglocorys cava70.5 (3.5)54.0 (4.7)41.7 (5.1)24.5 (6.0)  44.68 (<0.001)
CestodaBertiella sp.15.6 (2.8)15.0 (3.4)6.3 (2.5) 7.5 (3.7)7.52 (0.06)
TrematodaDicrocoeliidae sp. 0.6 (0.60)0.9 (0.90)0.0 (-)   1.9 (1.9)2.17 (0.54)
Table 5. Parasite egg counts (helminths) and infection intensity (protozoa) by site. Multiple samples from individuals included. Final column shows test results (SAS GENMOD adjusting for overdispersion) of differences in infection prevalence across sites. Aside from protozoa, intensity values here represent egg count intensity. (SE = standard error; IQR = interquartile range; dof = degrees of freedom)
  SonsoBusingiroKaniyo-PabidiKasokwa 
  Mean (SE)Median (IQR)Mean (SE)Median (IQR)Mean (SE)Median (IQR)Mean (SE)Median (IQR)GENMOD χ2, P (3 dof)
NematodaOesophagostomum10.09 (1.41)  4.0 (1.0–10.0)7.48 (1.56)2.0 (0.0–6.0)5.75 (1.51)2.0 (1.0–6.0)6.85 (1.81)1.0 (0.0–7.0)4.85, 0.051
 Necator1.30 (0.26)0.0 (0.0–1.0)1.50 (0.26)0.0 (0.0–2.0)1.11 (0.27)0.0 (0.0–1.0)1.62 (0.69)0.0 (0.0–1.0)1.11, 0.780
 Trichostrongylus1.24 (0.34)0.0 (0.0–1.0)2.83 (0.52)1.0 (0.0–3.0)1.50 (0.45)0.0 (0.0–1.0)1.64 (0.55)0.0 (0.0–1.0)7.52, 0.06
 Probstmayria0.20 (0.04)0.0 (0.0–0.0)0.38 (0.17)0.0 (0.0–0.0)0.08 (0.03)0.0 (0.0–0.0)0.08 (0.04)0.0 (0.0–0.0)8.86, 0.032
 Strongyloides1.13 (0.24)0.0 (0.0–1.0)2.93 (0.57)1.0 (0.0–3.0)0.45 (0.11)0.0 (0.0–1.0)3.64 (0.87)1.0 (0.0–3.0)38.1, <0.001
 Trichuris0.06 (0.05)0.0 (0.0–0.0)0.36 (0.19)0.0 (0.0–0.0)0.01 (0.01)0.0 (0.0–0.0)0.77 (0.74)0.0 (0.0–0.0)10.74, 0.0132
 Ascaris0.01 (0.01)0.0 (0.0–0.0)0.00 (0.00)0.0 (0.0–0.0)0.00 (0.00)0.0 (0.0–0.0)0.00 (0.00)0.0 (0.0–0.0)4.63, 0.20
 Unid. Strongyle0.37 (0.14)0.0 (0.0–0.0)0.15 (0.04)0.0 (0.0–0.0)0.15 (0.05)0.0 (0.0–0.0)0.66 (0.27)0.0 (0.0–0.0)7.60, 0.055
 Total Nematodes15.42 (1.62)  8.0 (3.0–17.0)18.0 (2.32)   9.0 (4.0–23.0)10.52 (1.91)  6.5 (3.5–11.5)19.38 (2.87) 12.0 (6.0–22.0)8.62, 0.039
ProtozoaUnidentified Protozoa149.42 (36.22)0.0 (0.0–0.0)90.24 (21.18)0.0 (0.0–0.0)126.72 (25.79) 0.0 (0.0–1.0)68.51 (30.40)0.0 (0.0–0.0)3.68, 0.30
 Troglodytella30.67 (6.21) 7.0 (3.0–20.0)42.54 (10.96) 5.0 (2.0–14.0)40.53 (12.54) 5.0 (2.0–16.0)23.38 (9.86)  4.0 (2.0–14.0)2.13, 0.55
 Balantidium coli0.04 (0.04)0.0 (0.0–0.0)0.00 (0.00)0.0 (0.0–0.0)0.00 (0.00)0.0 (0.0–0.0)0.00 (0.00)0.0 (0.0–0.0)38.5, <0.001
 Troglocorys cava23.47 (6.51) 2.0 (0.0–9.0)20.60 (7.78) 1.0 (0.0–5.0)2.69 (0.71)0.0 (0.0–2.0)1.25 (0.49)0.0 (0.0–0.0)15.9, <0.001
 Total ciliates54.17 (8.85) 12.00 (6.0–38.0)63.14 (13,26) 8.0 (3.0–29.0)43.22 (12.15) 8.0 (4.0 = 21.5)24.66 (9.82)   5.0 (2.0–18.0)4.83 (0.18)
CestodaBertiella2.35 (1.06)0.0 (0.0–0.0)2.37 (1.38)0.0 (0.0–0.0)0.34 (0.25)0.0 (0.0–0.0)1.91 (1.30)0.0 (0.0–0.0)3.22, 0.60
TrematodaDicrocoeliidae sp.0.01 (0.01)0.0 (0.0–0.0)0.01 (0.01)0.0 (0.0–0.0)0.0 (0.0)  0.0 (0.0–0.0)0.06 (0.06)0.0 (0.0–0.0)7.51, 0.051

Differences in disturbance

Levels of disturbance differed at each site, along different measures (Table 6). Kaniyo Pabidi had the least disturbance and Kasokwa had the greatest number and variety of disturbances. Sonso had the greatest number of trails. A Kruskal-Wallis test revealed significant differences in the number of trails (χ2[3] = 17.38, P = 0.001), tree stumps (χ2[3] = 75.16, P < 0.001), evidence of past logging sites (χ2[3] = 22.60, P < 0.001), evidence of current logging (χ2[3] = 30.40, P < 0.001), pole collection (χ2[3] = 29.64, P < 0.001), and charcoal pits (χ2[3] = 13.46, P = 0.004).

Table 6. Number of disturbances recorded in each plot at different sites. Site with greatest form of each disturbance is in bold
SiteStatisticTree stumpSnareCharcoal pitCurrent logging siteOld logging siteTrailPole cutting
KasokwaMean1.70.10.10.200.20.3
 Std. deviation1.80.50.20.400.40.4
SonsoMean1.2 0.02 0 0.030.10.40.2
 Std. deviation4.30.2 00.20.30.50.4
BusingiroMean0.60.1 00.2 0.030.10.3
 Std. deviation1.20.5 00.40.20.30.5
KPMean 0 0 0 0 00.2 0
 Std. deviation 0 0 0 0 00.4 0

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Thirteen different parasite species infected Budongo chimpanzees. Prevalence of some nematodes (Oesophagostomum) and ciliates (Troglodytella) reached over 80% (Table 4). These patterns are comparable to Gombe Stream National Park in Tanzania where results of the same analyses recovered 17 species of parasites with similarly high prevalence values for Oesophagostomum and Troglodytella (Gillespie et al. 2010).

Individuals that spent more time close to, or on, the ground had a higher parasite burden (Figure 2). This is the first study to illustrate substrate use as a factor influencing pathogen transmission risk in chimpanzees. Although the mechanism of transmission needs elucidation, it is reasonable to assume the soil may be contaminated with infective stages of parasites. Terrestrial individuals may have increased disease burden as a result of increased exposure. Repeated use of trails could further result in parasitic reinfection and cross infection. However, as both time on the ground and mean height were predictors of infection status (total number of infections, allowing for their interrelatedness), results indicate that there may be more than one route to infection (e.g., infection from other arboreal primates).

Chimpanzees at Sonso, the long-term research site, spent a greater percentage of time on the ground and were seen using human trails more often than chimpanzees at other sites (Figure 1). We hypothesize that increased ground use may result from the increased availability of trails and from continued exposure to researchers. Trail use may improve foraging efficiency, whereas researcher presence may reduce the risk of encountering predators on the ground (Boesch 1991). Thus, activities associated with research may alter the potential trade-offs between foraging and predation risk, encouraging ground use by habituated chimpanzees. Our hypothesis is supported by previous study documenting increasing ground use and vertical niche expansion in muriqui groups (Brachyteles hypoanthus) exposed to long-term research (Strier 2010). Perceptions of predator risk have also been found to be context-dependent in other species (Périquet et al. 2010).

Unfortunately, without a detailed camera trap study, we cannot determine “natural” levels of ground use in unhabituated chimpanzees. This may range along a spectrum. However, in areas with predators, we assume that chimpanzees would minimize time on the ground due to predation risk. In forests, leopards (Panthera pardus), the main predator of chimpanzees, often hunt from the ground (Zuberbuhler & Jenny 2002).

Other explanations for Sonso chimpanzee ground use are plausible, though not compelling. Ground use may reflect differences in season or maximum temperature (Takemoto 2004; Kosheleff & Anderson 2009). However, data were collected at sites with similar climates, during the same period of time (season and time of day). While microclimate may differ according to structural components of a landscape, temperature differences, and consequently ground use, should be most extreme in Kasokwa, a forest fragment with significant edges (Chen et al. 1999). Sonso chimpanzees may use the ground due to canopy gaps created by the sawmill. However, their home range extends beyond this area (Newton-Fisher 2003) and they spent the majority of time in closed or open tropical forest (Table 2). Again, Kasokwa should be the site most conducive to chimpanzee ground use as it lacks a continuous forest canopy.

For conservation purposes, it is critical to determine whether parasite infections obtained as a result of terrestrial activity result in reduced fitness. The strongyle nematodes have been associated with ulceration, anemia, protein malnutrition, weight loss, and death in primates (McClure & Guilloud 1971; DePaoli & Johnsen 1978). However, the pathology of parasites is variable. Some chimpanzees infected with Oesophagostomum sp. develop nodular oesophagostomosis without severe clinical signs (Krief et al. 2008). Others suffer from oesophagostomosis-associated morbidity and mortality (Huffman et al. 1997; Gillespie et al. 2010).

Predicting the effect of human disturbance on parasitism is complicated by individual variation in sex, age, nutrition, and stress (Gillespie 2006). This study focused on behavioral changes in opportunities for parasite transmission. We demonstrated that different rates of parasitism correspond with a tendency to stay close to the ground. We believe this tendency may be promoted by long-term researcher presence and the availability of man-made trails. Results will have to be replicated. Nevertheless, these results contribute to debates surrounding risks of researcher activities. A variety of measures have been implemented at research sites to limit the potential impact of researchers on wildlife. These include reducing the number of researchers, increasing the viewing distance, and improving sanitation (Boesch 2008). Similar rules existed at Sonso. Nevertheless, our results indicate that the presence of researchers themselves and the tools used in research (i.e., trails) may influence the parasite ecology of wildlife. Prior to conducting research of endangered species, it is critical to evaluate the conservation risks facing study populations (disease, predation, hunting). If disease outbreaks are a threat at a particular site, it may be prudent to minimize trail creation and maximize use of remote methods of monitoring (e.g., camera traps and drones). Data on parasite burdens should be collected systematically to monitor changes in the health of study populations over time (Bakuza & Nkwengulila 2009).

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Research permission was granted by the Uganda Wildlife Authority, the Ugandan National Council of Science and Technology, and the Office of the President of Uganda. We thank E. Canfield and N. Hauser for assistance with laboratory analyses, and J. Cilliers for helpful comments during writing. Research was funded by Oxford University, the Cleveland Metroparks Zoo, and Emory University.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References